PROJECT SUMMARY/ABSTRACT Malignant pleural mesothelioma (MPM) is a rare cancer with poor prognosis. There have been no FDA- approved therapies for MPM patients since 2003. Patients with MPM who present with high levels of cytotoxic tumor infiltrating lymphocytes (TILs) have better survival. To promote TILs in MPM, we developed an adoptive T-cell therapy using chimeric antigen receptors (CARs). We have developed and translated mesothelin (MSLN)?targeted CAR T cells to phase I clinical trials (NCT02414269 and NCT02792114). MSLN is a cell- surface antigen highly expressed in MPM, with very low expression in normal tissues. In our two clinical trials, no on-target, off-tumor toxicities were noted among 28 patients treated so far. CAR T cells are administered intrapleurally in NCT02414269, on the basis of our published data (Sci Transl Med 2014) that established regionally administered CAR T cells potentiate antitumor efficacy by augmenting CD4 T-cell helper function. Beyond safety, promising antitumor efficacy has been observed in our phase I trial of intrapleural CAR T cells. Following administration of a low dose of CAR T cells, tumor cells upregulate PD-L1/L2 and inhibit T cells via binding to PD-1. We have shown that administration of anti-PD-1 agents can overcome tumor-mediated adaptive resistance and promote CAR T-cell functional persistence (J Clin Inv 2016). Supported by these data, we treated 11 of 18 MPM patients with pembrolizumab, an anti-PD-1 agent, and noted no adverse events, and enhanced persistence of CAR T cells and antitumor efficacy. Building on these strong data, we propose an investigator-initiated phase I/II clinical trial combining CAR T-cell therapy with pembrolizumab. We hypothesize that cancer antigen-targeted CAR T cells can promote TILs and that anti-PD-1 agent can combat adaptive resistance by reversing exhaustion of both CAR and endogenous T cells. We will determine the safety of adoptive transfer of genetically modified, autologous, MSLN-targeted T cells into the pleural cavity of MPM patients (with previous administration of cyclophosphamide), followed by treatment with pembrolizumab until tumor eradication or toxicity (Aim 1). The primary endpoint is to determine the safety and maximum tolerated dose (MTD) of MSLN-directed CAR T cells in combination with pembrolizumab (phase I) and to estimate the response rate of the combination therapy (phase II). We will analyze serially collected pleural fluid, tumor biopsy specimens, and peripheral blood to assess the ability of PD-1 blockade to reverse CAR T-cell exhaustion (Aim 2) and activate antitumor endogenous immunity (Aim 3), compared with treatment with pembrolizumab alone. Investigating regional versus systemic endogenous and CAR T-cell immunity following activation with anti-PD-1 agent, compared with anti-PD1 agent alone, is mechanistic, innovative, and translational. The ensuing results are directly applicable to 150,000 patients with pleural metastatic tumors (from MSLN+ve lung and breast cancers) and extendable to more than 2 million patients with MSLN-expressing solid tumors per year in the U.S. alone.